Multi-Sensor Arrays for Online Monitoring of Cell Dynamics in in vitro Studies with Choroid Plexus Epithelial Cells

Sensors 2012, 12, 1383-1397; doi:10.3390/s120201383 sensors ISSN 1424-8220 www.mdpi.com/journal/sensors Article Multi-Sensor Arrays for Online Monitoring of Cell Dynamics in in vitro Studies with Choroid Plexus Epithelial Cells Pedro Mestres-Ventura 1,2,*, Andrea Morguet 1 and Soledad García Gómez de las Heras 2 1 Department of Anatomy and Cell Biology, Saarland University, University Hospital, Bldg. 61, 66421 Homburg, Saar, Germany; E-Mail: wm[email protected] 2 Department of Human Histology and Pathology, Faculty for Health Sciences, University Rey Juan Carlos I, Av. de Atenas s/n, E-28922 Alcorcón, Madrid, Spain; E-Mail: soledad.ga[email protected] * Author to whom correspondence should be addressed; E-Mail: [email protected]; Tel.: +34-91-488-8622; Fax: +34-91-488-8955. Received: 1 December 2011; in revised form: 13 January 2012 / Accepted: 17 January 2012 / Published: 1 February 2012 Abstract: Sensors and multi-sensor arrays are the basis of new technologies for the non-label monitoring of cell activity. In this paper we show that choroid plexus cells can be cultured on silicon chips and that sensors register in real time changes in their activity, constituting an interesting experimental paradigm for cell biology and medical research. To validate the signals recorded (metabolism = peri-cellular acidification, oxygen consumption = respiration; impedance = adhesion, cell shape and motility) we performed experiments with compounds that act in a well-known way on cells, influencing these parameters. Our in vitro model demonstrates the advantages of multi-sensor arrays in assessment and experimental characterization of dynamic cellular events—in this case in choroid plexus functions, however with applicability to other cell types as well. Keywords: sensor arrays; peri-cellular acidosis; cell respiration; cell adhesion; impedance; cell cultures; electron microscopy; choroid plexus; chick embryo 1. Introduction Electrical and electrochemical sensors have enabled the development of technologies, which are very useful in obtaining data on cellular activities with minimal external intervention. These techniques can be applied without any labeling of cells, thus avoiding the possible alteration of cells OPEN ACCESS Sensors 2012, 12 1384 under investigation. Furthermore, these techniques have paved the way for online observation, enabling continuous, uninterrupted execution of the experiment. These developments represent a considerable improvement on the more traditional so-called “end-point” techniques [1]. Pioneer devices were equipped with sensors detecting changes in cell adhesion, cell motility and cell shape, based on impedance measurements [2]. Another generation of sensors was the so-called light-addressed potentiometric sensors (LAPS), able to detect changes in peri-cellular acidosis, a phenomenon which is closely related to the state of cell metabolism [3]. Moreover, to ensure the function and vitality of the cells, these devices were equipped not only with a fluidic system that ensured the supply of culture medium, but also with systems for regulating temperature, humidity and gas atmosphere in the modules where the cells were incubated. A common feature of the apparatus of this first generation was that they were mono-parametric [4]. During the past 20 years these devices have found numerous applications in biology and medicine, for instance, in the study of ligand-receptor interactions, in the analysis of the action of chemicals and drugs on specific cell targets, in the assessment of the chemosensitivity of cancer cells, and many more [5–7]. These sensors were followed by a new generation of sensors characterized by the presence on the silicon chip of several sensor types, the so-called multi-sensor arrays, with which many parameters can be measured simultaneously: impedance (adhesion, cell shape), acidification or peri-cellular acidosis (metabolism) and oxygen consumption (respiration) [8–10]. From the very beginning, these new systems were developed with a view to their possible applications in research areas such as the chemosensitivity of tumor cells and drug screening. More recently, examples of successful applications in toxicology, pharmacology, and even in the field of ecology can be found in relevant literature [11–14]. With the application of multi-sensor technology arose the necessity for the development of methods of preparation of cells and tissues suitable for the investigation of such specimens in these new devices. At first, the use of permanent cell lines was imperative, because they are easier to handle in the sensor devices. But in the meantime, considerable experience has been acquired with primary and organotypic cultures, as well as tissue slices [15]. This paper describes the preparation and management of choroid plexus epithelial cells (CPC) on multi-sensor arrays as an in vitro cell model and presents results of how these cells react when treated with compounds influencing their metabolism and transport functions. The significance of choroid plexus lies in the fact that it is a major source of cerebrospinal fluid, which circulates through the brain ventricles and the subarachnoid space [16]. Because of the complex network of interactions in the whole body, appropriate cellular models under axenic conditions are required to understand the molecular basis of specific mechanisms of the choroid plexus epithelial cells, since they are the main cell element of this organ. These requirements were taken into account in the experimental approach described here with a resulting high relevance for cell biology in general and medical research in particular (neurology, neurosurgery, pharmacology, etc.). 2. Experimental Section 2.1. Cell Cultures Choroid plexus of chicken embryo 14 ED embryos (20–30 embryos pro assay) were dissected and collected in culture medium (DMEM/F12 Gibco, supplemented with penicillin and streptomycin, Sensors 2012, 12 1385 L-glutamine, insulin, EGF, hydrocortisone, NEA, transferrin, 10% FCS). After dissection, the choroid plexus were dissociated with trypsin + EDTA (Gibco) in a Petri dish and using a magnetic stirrer at 37 °C in a CO2 incubator (Heraeus). Subsequently, the cell suspension was passed through a cell strainer (BD) and rinsed with medium with 10% FCS (to inactivate trypsin). The resulting cell suspension was centrifuged at 2,400 rpm for 4 min (cyto-centrifuge Sartorius), the supernatant discarded and the pellet re-suspended in fresh medium and the cells counted with Trypan Blue (Sigma-Aldrich). Subsequently, the chips were inoculated with 1.5 or 3 × 105 cells each. Parallel to this, well plates with Thermanox slides (Nunc) were also seeded with a similar cell suspension (2 × 106 pro well). Although the chips are made of silicon, a material suitable for cell cultures, we coated their surfaces with collagen I, which considerably improves attachment conditions for the choroid plexus cells [17]. A solution of collagen I (20 µg/mL in 0.02 N acetic acid) was applied to the surfaces of the chips, which were then incubated at 37 °C for 1 h. Subsequently, they were washed with culture medium and used immediately or refrigerator-stored. Thermanox slides were coated following same protocol. The cells were first cultured without cys-arabinoside (Sigma-Aldrich) on both chips and Thermanox slides. After 24 h the medium was replaced by fresh medium containing cys-arabinoside, a substance able to suppress the growth of fibroblasts and endothelial cells, but which does not affect the epithelial cells of the choroid plexus. After a culture period of 24 h or longer (until 70–80% of the chip surface was covered with choroid plexus cells) the chips were transferred to the sensor device. The measurements in the sensor device were conducted with a medium (low buffering capacity), which did not contain cys-arabinoside. The measurement times in the sensor unit ranged from 24 to 72 h, or even longer according to experiment requirements. The growth of cells on Thermanox plates was regularly monitored using an inverted microscope (Olympus) with phase contrast. As, in this case, the chips were opaque, the examination was performed by reflected light with a stereomicroscope (Wild, Heerbrugg, Switzerland). 2.2. Sensor Device The Bionas 2500® analyzing system (Bionas GmbH, Rostock, Germany) was used in these studies. A detailed description of this system has been published previously [13]. Briefly, the device is composed of six modules, which are connected with a pump and control units. The culture media, with or without drugs, first flow to and then through each module containing the sensor chips and are then directed to a waste collector. The reference electrode is placed in the fluidic pathway behind the sensor modules. The sensor chip (metabolic chip Bionas Discovery™ SC1000) used here, contains a sensor array composed of: (a) two Clark-type sensors, (b) five ISFETs (ionic-sensitive field-effect transistors) and (c) one IDES (inter-digitized electrode sensors). Such sensors allow simultaneous determination of cellular oxygen consumption (a), peri-cellular acidification (b) and cell impedance (c) in on-line regime and without any labeling. Before the sensor analyzer is used, the tubes of the fluidic system have to be disinfected by perfusing them with 70% ethanol, followed by PBS and a final washing with culture medium, in our case DMEM/F12 (see above). The sensor chips should also be disinfected according to a similar protocol as above. The sensor chips with CPC cells (see above) were placed into the modules and maintained in the Bionas device at 37 °C. Sensors 2012, 12 1386 The following operational parameters of the Bionas Analyzing System 2500® were selected. The flow rate was 56 µL/min with an alternate cycle of “go” and “stop” phases of 4 min each. During the “go” phase choroid plexus cells attached to the chip received fresh medium with or without compounds (see below). During the “stop” phase oxygen consumption and extracellular acidification were continuously measured in the supernatant. Changes in the rates of these parameters were detected for each single stop phase, representing the metabolic activity of the cells at that time and are given as single data points in the diagrams. The presence of cells in contact with IDES was detected and expressed in terms of cell impedance. These measurements provided information on the extent of cell adhesion to the substrate precisely at the IDES areas and were independent of pump activity. However, in consideration of the influence of interferences, the data point values during the “stop” phase were averaged. All data obtained have been presented normalized and were processed with Origin 8.0. 2.3. Compounds To induce interpretable changes in metabolism and behavior of choroid plexus cells the following compounds were selected: (a) Sodium fluoride (NaF, Merck) that dose-dependently inhibits glycolysis and anaerobic metabolism; in addition, direct effects of NaF on mitochondria have been described (see below) (b) Potassium cyanide (KCN, Riedel-de-Haen) that inhibits cellular respiratory processes, especially at the level of mitochondria, (c) Forskolin (generously provided by Prof. Dr. M. Diener, University of Giessen, Germany), an agonist of adenylate cyclase, an enzyme that stimulates and regulates the synthesis of cAMP, (d) Acetazolamide (Sigma-Aldrich) that inhibits carbonic anhydrase, a key enzyme in water transport processes in the choroid plexus. These compounds were implemented in the following concentrations: NaF 20 mM, KCN 5 mM and 10 mM, forskolin 5, 10 and 25 µM, acetazolamide 10 µM, 100 µM, 1 mM and 2 mM. The compounds were applied during periods of time ranging from 1 to 24 h, followed by a variable period, during which the cells received culture medium containing no compound, with the aim of detecting possible late effects and estimating the degree of cell recovery after a certain treatment. At the end of each experiment, the chips were examined and photographed under a stereoscopic microscope, and then fixed chemically for electron microscopy. The Bionas device has six modules that were distributed as follows: two modules were destined for the controls, two for compound 1 and the remaining two for compound 2. This procedure facilitates comparative analysis. At least four chips per compound and concentration were measured and evaluated. Chips which created disturbances during experiments were excluded. The experiments were performed (per compound and concentration) on at least two different days. 2.4. Electron Microscopy Chips and Thermanox slides were fixed in 3% glutaraldehyde (Polysciences Inc.) in PBS. For scanning electron microscopy in a FEI 30XL ESEMTM at 10 kV, specimens were post-fixed with Sensors 2012, 12 1387 osmium tetroxide, dehydrated in a series of alcohols, critical point dried and sputtered with platinum [18,19]. Transmission electron microscopy followed a protocol described elsewhere [20,21]. Briefly, fixation was performed with 3% glutaraldehyde in PBS, post-fixation with osmium tetroxide, dehydration in a series of alcohols and embedding in Embed 812 (Electron Microscopy Sciences, USA). Epoxy resin mixture was poured on the cells and polymerized at 60 °C. The polymerized resin layer with the cells can be easily removed from the silicon chip. Ultrathin sections (80 nm thick) were stained with uranyl acetate (Merck) and lead citrate (EMS, USA) and examined in a FEI TecnaiTM Biotwin 120 transmission electron microscope at 60 kV. 3. Results and Discussion 3.1. Cultures of Choroid Plexus Epithelial Cells (CPC) The preparation of the choroid plexus of the chick embryo was performed—with slight modifications—according to a protocol already described [22]. Following this protocol, we have achieved the formation of cell monolayers composed almost exclusively of choroid plexus epithelial cell (Figure 1). This was made possible by the addition of cys-arabinoside, an inhibitor of DNA synthesis, to the culture medium, [22]. The CPCs have a nucleotide transport system unable to uptake cys-arabinoside; therefore growth and morphology of these cells remain unchanged. In contrast, the other cells of plexus (fibroblasts, endothelia) take up the cytostatic, which prevents their proliferation and growth. The result is pure cultures of CPC. Figure 1. Phase contrast of choroid plexus cells grown on Thermanox for 4 days. The cells form a monolayer displaying a polygonal shape. The culture was previously treated with cys-arabinoside to eliminate cell contamination. 20X objective. When the cell suspension was applied to either chips or Thermanox slides, the cells attached at an early stage during the first 24 h, showing a variable shape, consistent with the process of attachment to substrate. After 2 days cultivation the cells began to aggregate with their congeners, constituting the monolayer 2 days later (Figure 1). Comparative studies with other coatings (laminin, fibronectin and collagen IV) have shown that CPCs attach faster on collagen I-coated surfaces, a circumstance beneficial to our studies with the sensors, as we can start the measurements earlier. Sensors 2012, 12 1388 Electron microscopy revealed the surface ultrastructure of these monolayers and confirmed their specific cell composition (Figure 2). The cell borders were marked by numerous thin microvilli, giving the image a mosaic-like appearance. In the center of the cells, apical pole microvilli and cilia, typical attributes of these cells lost during enzymatic dissociation, reappeared. Transmission electron microscopy confirmed these diagnoses, and also enabled determination of the participation of cytoskeleton and cell membrane, forming intercellular contacts and anchorages to substrate (not shown). Cells grown on silicon chips or Thermanox displayed exactly the same patterns. Figure 2. Scanning electron microscopy of CPCs growing on the sensor chip. The cells have numerous microvilli at the apical pole and show a marked relief in the central part that corresponds to the nucleus. Before the silicon chips with the cells were placed into the sensor device, every chip was examined under a stereomicroscope to estimate the quality of the monolayers and their topographic relationship with the sensors (Figure 3). Figure 3. Scanning electron microscope image shows groups of cells arranged in monolayer growing on the chip. The CPCs can be seen as a grey veil covering the major part of the visible chip surface. S 3 I c m s ( d a t h S ensors 201 3 .2. Measu r CPC res p SFET, Cla r c urve dropp m etabolism s tabilizing a Fi g ur e graph ) stron g With the approxima t d etached (F i a gain, reac h h e cells ha d Fi g ur e decre a 2, 12 r ements un d p onded to N r k). The m e ed immedi a recovered v a few hours e 4. Onlin e ) . Note the g activation. presence o f t ely 40%), i gure 5). W h ing the init d re-adhere d e 5. Online a sed very q u d er Experim N aF with c h e tabolis m m a tely to 80 % v ery quickl y later and b e e monitori n biphasic c Start and S f NaF at th e which m e W hen the m e ial values p d . monitorin g u ickly and r e ental Con d h anges in a m easuremen t % (Figure 4 ) y (upward c u e having lik e n g of meta b c urve indic a S top: Begin n e module, t h e ans that a e dium was r p rior to the a g of cell ad h e covery tak e d itions a ctivity det e t s (peri-cel l ) . When th e u rve), reac h e the contro b olism of C a ting a str o n ing and e n h e cells rea c a dhesion w r eplaced w i a pplication h esion of C P e s a long ti m e ctable wit h l ular acidifi e medium w h ing values h l. C PCs expos o ng inhibit i n d of comp o c ted with a n w as reduce d i th fresh m e of fluoride P Cs expos e m e to reach h all three t cation or a c w as replace d h igh above ed to fluor i i on, follow o und applic a n immediat e d and con s e dium wit h . These me a e d to fluori d values simi l t ypes of se n c idosis) sh o d with fresh the initial v i de (detail s ed by a si m a tion respe c e decrease i s equently c h out NaF, t h a surements d e. The im p lar to the c o 13 8 n sors (IDE S o wed that t h medium, t h v alue (160 % s in the m ilarly c tively. i n impedan c c ells beca m h e curve ro s indicate th p edance o ntrols. 8 9 S , h e h e % ), c e m e s e at S r u ( m u ( A c S ensors 201 NaF dis c r ecovered a u nstable pr o Fi g ur e This p CPC ex p over 140% ) m edium, re v u nchanged u Figure 8). B A fter remo v c ontrols. T h Fi g ur e wellk 2, 12 c retely inh i fter the re m o file (Figur e e 6. Online p arameter d e p osed to po ) , i.e., with v erted with i u nder these B y contras t v al of the c h is applies f o e 7. Onli n k nown effe c i bited cell u m oval of N e 6). monitorin g e creases, bu t tassium cy a higher per i i n a few ho u conditions , t , cellular r c yanide, th e o r both use d n e monitor i c t of cyanid e u lar respir a N aF, still re m g of respira t t does not r e a nide (KC N i -cellular a c u rs to level s , indicating espiration u e cells rec o d concentra i ng of res p e on oxyge n a tion (curv e m ained be l t ion (oxyge n e cover com p N ) respond e c idification , s similar to that no rel e u nderwent a o vered wit h tions of 5 a n p iration of n consumpt i e down b y l ow the co n n consump t p letely (co m e d with a r which, aft e control lev e e vant chan g a significa n h in a few h n d 10 mM o CPCs ex p i on, with r e y 30%), w n trol levels t ion) of CP C m pare with t r emarkable e r the rem o e ls (Figure 7 g es in cell a n t drop of a h ours, reac h o f KCN . p osed to c e covery wit h w hich, alth o and displ a C s exposed t he control) . increase i n o val of cya n 7 ). Impeda n a dhesion ha d a pprox. 60 % h ing levels c yanide. N o h in a few h o 13 9 o ugh it lat e a yed a rath e to NaF. . n metabolis m n ide from t h n ce remain e d taken pla c % (Figure 9 close to t h o te the o urs. 9 0 e r e r m h e e d c e 9 ). h e S g c s t h d t h o w d f S ensors 201 Fi g ur e as Fi g with c Fi g ur e type a The met a g lycolysis [ 2 c ell [3,6,15] s tress, inclu d h at accom p d etachment h e changes o f impairm e The exp e w ith the bl o d oes not ap p f rom the gl y 2, 12 e 8. Online g ure 7). No t yanide. e 9. In co n s experime n a bolic chan g 2 3], resulti n . Moreove r d ing chang e p anied thes e of the cell s in respirat o e nt of the st a e riments w i o cking of r e p ear to be l e y colysis, w h monitorin g t able activa t n trast, adhe s n t Figure 7 ) g es triggere n g in a redu r , it has bee n e s in cell p r e processes s [26]. Suc h o ry parame t a bility of t h i th KCN ar e e spiration c o e thal (unde r h ich appea r g of metabo t ion of me t s ion of CP C ) . d b y NaF a r ction in the n well-doc u r oteins, mit o diminishe d h events a r t ers could b h eir membr a e interesti n o uld have a r the condi t r ed strongl y lism of CP C t abolism (a c C s (imped a r e explaine d extrusion o u mented th a o chondria a d membran e r e consiste n b e due to th e a nes throug h n g, as the i n a compens a t ions used h y activated i C s exposed c idification ) a nce) does n d by the in h o f protons ( p a t NaF can i n d cell nuc l e fluidity, w n t with our e effect of N h lipid pero n crease in g a tory signif i h e r e) becau s i n these ex p to cyanide ) observed n ot show a n h ibition of e p eri-cellul a i nduce apo p l eus [24,25 ] w ith the app e impedance N aF on the xidation [2 7 g lycolysis o i cance. Th e s e the ATP p eriments ( (same exp e in all expe r n y change s e nolase, a k e a r acidificat i p tosis throu g ] . The lipid e arance of b measurem e mitochond r 7 ]. o bserved si m e blocking o needed ca n increase o f 13 9 e riment r iments s (same e y enzyme i i on) from t h g h metabol i peroxidati o b lebbing a n e nts. Finall y r ia as a res u m ultaneous l o f respirati o n be provid e f peri-cellul a 9 1 i n h e i c o n n d y , u lt l y o n e d a r